Development of the human fetal kidney begins at approximately

Transcription

1 Suppression of Ureteric Bud Apoptosis Rescues Nephron Endowment and Adult Renal Function in Pax2 Mutant Mice Alison Dziarmaga,* Michael Eccles, and Paul Goodyer* Departments of *Human Genetics and Pediatrics, Montreal Children s Hospital, McGill University, Montreal, Québec, Canada; and Department of Pathology, University of Otago, Dunedin, New Zealand The molecular mechanisms that set congenital nephron number are unknown. However, humans with modest suboptimal nephron number may be at increased risk for essential hypertension, and those with more severe nephron deficits at birth may develop progressive renal insufficiency. A model of branching morphogenesis during fetal kidney development in which the extent of ureteric bud arborization is dependent on suppression of programmed cell death has been proposed. This study shows that the increased apoptosis and reduced ureteric bud branching of heterozygous Pax2 mutant mice is associated with 40% decrease in nephron number at birth. This leads to postnatal glomerular hypertrophy and long-term renal insufficiency in the absence of glomerulosclerosis. To determine whether restoration of antiapoptotic factors alone is sufficient to rescue the nephron deficit in these mice, a BCL2 transgene that is under the control of the PAX2 promoter was targeted to the ureteric bud. The transgene suppressed programmed cell death in the ureteric bud lineage, increased nephron number to 90% of that of wild-type littermates at birth, and normalized renal function at 1 yr. These observations lend strong support to the hypothesis that factors that control ureteric bud apoptosis are powerful determinants of congenital nephron endowment. J Am Soc Nephrol 17: , doi: /ASN Development of the human fetal kidney begins at approximately 4 to 5 wk of gestation, when the ureteric bud (UB) emerges from the wall of the nephric duct. Focused trophic signals from the neighboring metanephric mesenchyme guide growth of the UB outward. As it penetrates lateral mesenchyme, the UB begins to arborize and induce nephron formation at the tip of each branch. Programs that optimize the rate of UB branching are critical because new nephron formation comes to an end approximately 1 mo before birth in humans. Therefore, the extent of branching during fetal development determines nephron endowment for life. Nephron endowment seems to vary widely among normal individuals. From autopsy studies in which formal nephron counts were performed, it seems that normal nephron number ranges from 300,000 to 1 million per kidney (1). Although this was once considered to be normal variance without clinical significance, Brenner and colleagues (2 4) argued that suboptimal nephron number at birth may increase the risk for development of hypertension and/or increased susceptibility to acquired renal insults later in life. Rat strains with low nephron number have increased BP and increased susceptibility to renal injury (2 4). Cullen-McEwen et al. (5) recently showed that the loss of one glial cell-derived neurotrophic factor (Gdnf) allele results in 30% fewer glomeruli in young mice. As these mice Received October 17, Accepted March 7, Published online ahead of print. Publication date available at Address correspondence to: Dr. Paul Goodyer, Montreal Children s Hospital Research Institute, 4060 Saint Catherine West, Room PT-413/1, Montreal, Quebec, H3Z 2Z3 Canada. Phone: , ext ; Fax: ; paul.goodyer@muhc.mcgill.ca age, the glomeruli undergo compensatory hypertrophy, enlarging to approximately twice the size of glomeruli in control animals. Although GFR is normalized by glomerular hypertrophy, Gdnf heterozygotes develop hypertension in adult life (5). Similarly, Keller et al. (6) observed in an autopsy study that humans with essential hypertension had approximately half the number of glomeruli per kidney compared with matched normotensive control subjects. As with Gdnf mutant mice, each glomerulus was approximately twice the size of controls, and the prevalence of glomerulosclerosis was increased. The genetic programs that regulate nephron number are largely unknown. However, it was reported in 1995 that an autosomal dominant form of renal hypoplasia and ocular colobomas (renal coloboma syndrome) is attributable to heterozygous mutations of the PAX2 gene (7,8). PAX2 belongs to the nine-member family of paired box transcription factors and is highly expressed in the UB as it undergoes branching morphogenesis (9). It is interesting that reevaluation of some patients with the diagnosis of oligomeganephronia also showed some PAX2 mutations (10). In this syndrome, nephron structure seemed normal except that glomeruli were reduced in number and were strikingly hypertrophied (10). This report provided one of the first clues that PAX2 mutations cause a pure nephron deficit. In 1996, Favor et al. (11) reported a mouse strain with a spontaneous frameshift mutation of the Pax2 gene that permitted detailed study of renal phenotype during development. Fetal heterozygous mice were found to have kidney hypoplasia associated with increased apoptosis of UB cells and reduced number of UB branches (12). Previously, we used the PAX2 promoter to drive targeted expression of a proapoptotic (Bax) gene in normal UB cells and noted UB branching defects that Copyright 2006 by the American Society of Nephrology ISSN: /

2 J Am Soc Nephrol 17: , 2006 Targeted BCL2 Expression Rescues Renal Hypoplasia 1569 were comparable to Pax2 mutants (13). We proposed a model in which the rate of UB arborization is influenced by the balance of factors that determine UB susceptibility to programmed cell death (13). If this model is correct, then the regulation of UB cell survival should be a powerful determinant of congenital nephron endowment and eventual renal function. In this study, we show that heightened UB apoptosis and reduced branching in heterozygous Pax2 1Neu mice are associated with suboptimal nephron number at birth. Systematic counts of glomeruli show that Pax2 1Neu newborns have 40% fewer nephrons than do wild-type littermates. Furthermore, we show that this leads to postnatal glomerular hypertrophy and increased serum cystatin levels at 1 yr of age. Targeted expression of an antiapoptotic BCL2 transgene to the UB of Pax2 1Neu mutant mice reverses UB apoptosis, restores the rate of UB branching, normalizes congenital nephron number, and normalizes serum cystatin levels at 1 yr. Materials and Methods GFP-Pax2 1Neu Mice Hoxb7-GFP mice, provided by Frank Costantini (Columbia University, New York), were crossed with Pax2 1Neu ( / ) mutant mice. At 13.5 d of gestation, fetal embryos were removed, and kidneys were microdissected and photographed under fluorescence light. The hind limb was used for genotyping. RNA Isolation and Bcl2 Real-Time Reverse Transcription PCR RNA was isolated from embryonic day 15 (E15; four litters) and postnatal day 1 (P1; three litters) kidneys from wild-type and Pax2 1Neu ( / ) mutant mice. Kidneys were microdissected and placed in RNA- Later (Ambion, Austin, TX) at 4 C overnight; six to 16 samples were analyzed for each group. Each kidney was homogenized in Tri reagent (InVitrogen, Burlington, ON, Canada) and incubated at room temperature for 5 min. Twenty percent of total volume of chloroform was added, samples were mixed vigorously, incubated at room temperature for 10 min or until separate layers formed, and centrifuged at 12,000 rpm for 15 min at 4 C. The aqueous layer (top) was removed, and an equal volume of 70% ethanol was added. Samples then were processed with the RNeasy kit (Qiagen, Mississauga, ON, Canada) as per the manufacturer s recommendations. Samples were resuspended in 30 l of RNase-free water. cdna was synthesized from 500 ng of starting RNA using random primers and superscript III (Invitrogen, Burlington, ON, Canada) as per the manufacturer s recommendation. One microliter of cdna subsequently was used for quantitative real-time PCR with Sybr Green and ROX on an ABI prism 7000 machine. PCR conditions were as follows: 2 min of incubation at 95 C, followed by 40 thermal cycles of 95 C for 30 s, 60 C for 30 s, and 72 C for 30 s. Results were standardized with a housekeeping gene -2 microglobulin, and the comparative C T method was used for relative quantification. Primers were designed to span introns and their sequences are as follows: Bcl2 F, 5 -AGTACCTGAACCGGCATCTG-3 ; Bcl2 R, 5 -GCTGAGCA- GGGTCTTCAGAG-3 ; B2m F, 5 -TGCAGAGTTAAGCATGCCAG- TATGG-3 ; and B2m R, 5 -TGATGCTTGATCACATGTCTCG-3. Amplicons for BCL2 and B2M were 134 and 75 bp, respectively. Melt curves for each amplicon showed a single peak, indicating absence of primer dimerization or nonspecific PCR products. Each sample was run in duplicate. PAX2 Promoter-BCL2 Transgene Construction and Generation of Transgenic Animals A 4.0-kb human PAX2 promoter (AF515729) was ligated into a modified IRES vector as described previously (13). An EcoRV/SpeI fragment of human BCL2 cdna was ligated downstream of the promoter, using an end-filled NotI site and SpeI site in the IRES vector. XhoI linearized plasmid from the PAX2 promoter-bcl2 construct was diluted to 2 ng/ l and microinjected into the pronuclei of C57B/6J C3H fertilized ova. These eggs were injected into pseudopregnant CD1 female mice. Transgenic mice were identified by PCR amplification of the IRES region in DNA that was obtained from tails as described previously (13). Mice were backcrossed for five generations into a C3H background before analysis. Transgenic animals were fertile and viable. All mice were generated and maintained with approval from the Montreal Children s Hospital Research Institute Animal Ethics Committee, protocol #4242. BCL2 Immunohistochemistry E15.5 embryos were fixed overnight in 4% paraformaldehyde, dehydrated, and embedded in paraffin. Serial 7- m sections of the embryo were deparaffinized, rehydrated, and boiled twice in 10 mm citrate buffer (ph 7) for 5 min. Endogenous peroxidase activity was quenched in 3% H 2 O 2 (in methanol) for 5 min at room temperature. After a 30-min incubation with blocking horse serum, sections were incubated with primary antibody rabbit polyclonal anti-bcl2 antibody (Santa Cruz Biotechnologies, Santa Cruz, CA) at a 1:100 dilution overnight at 4 C. Vectastain ABC Universal Kit (Vector Laboratories, Burlingame, CA) was used as described by the manufacturer, followed by incubation with DAB (Vector Laboratories). Sections were counterstained with Mayer s hematoxylin solution (Sigma-Aldrich Canada Ltd, Oakville, ON, Canada), dehydrated, and mounted with Permount (Fisher Scientific, Pittsburgh, PA). Terminal Deoxynucleotidyl-Transferase Mediated dutp Nick-End Labeling Staining To determine the number of cells that underwent apoptosis in wildtype, Pax2 1Neu and Pax2 1Neu Bcl2, we performed terminal deoxynucleotidyl-transferase mediated dutp nick-end labeling (TUNEL; Roche Diagnostics, Laval, QC, Canada) as described previously on 7- m P1 sections (12). Sections were counterstained with hematoxylin solution. Photographs were taken at 10 objective on a Zeiss microscope. Maximal Kidney Cross-Sectional Area Determination Maximal cross-sectional area was determined for wild-type, transgenic, and Pax2 1Neu animals. Kidneys were dissected from P1 mice, fixed overnight in 4% paraformaldehyde, dehydrated, and embedded in paraffin. Serial 7- m sections were stained with hematoxylin and eosin. Every fourth section was assessed to identify the largest crosssection for each kidney. Mean cross-sectional area was measured with Spot software (Diagnostic Instruments, Sterling Heights, MI) on a Zeiss microscope for the animals in each group (wild-type n 12, Pax2 1Neu n 9, Pax2 1Neu BCL2, n 8). Nephron Counting Nephron number was assessed in P1 mouse kidney in two ways. First, glomeruli were counted in the largest sagittal cross-section from each kidney. The number of glomeruli per maximal cross-section was taken as a reflection of total kidney nephron number. Second, nephron number per kidney was estimated with the physical dissector/fractionator method in kidneys from four newborn (P1) pups from each

3 1570 Journal of the American Society of Nephrology J Am Soc Nephrol 17: , 2006 group (14,15). Every fourth and eighth serial sagittal section (7 m) were paired for analysis. Therefore, the distance between paired sections was 28 m; the distance between one section pair and the next was 112 m. Glomeruli were counted under a grid system; every second square was counted. When a glomerulus was present in one section but not in its paired section, it was counted. When Q is the actual number of glomeruli counted per kidney, N glom 112/28 2/1 1/2 Q. Cystatin C Measurements at 1 Yr of Age Adult mice were killed at 10 to 13 mo of age, and 300 l of blood was collected by cardiac puncture. Blood was centrifuged at 10,500 rpm for 10 min at room temperature in capillary blood collection tubes (Sarstedt, Numbrecht, Germany). After centrifugation, the serum was removed and stored at 20 C. Serum cystatin C was measured by nephelometry in wild-type (n 12), Pax2 1Neu (n 7), and Pax2 1Neu BCL2 (n 7) mice (16). Glomerular Hypertrophy and Histology at 1 Yr of Age To estimate glomerular volume, we calculated mean glomerular diameter from the largest 20 of 50 glomeruli in each sagittal section, presuming that the smaller glomerular cross-sections reflect tangential cuts. Four sections from each kidney (sections 40, 90, 140, and 190) were assessed to ensure that the same glomerulus was not measured twice. Therefore, the mean of the 80 largest glomeruli per kidney were used to calculate mean glomerular volume for wild-type (n 4), Pax2 1Neu (n 5), and Pax2 1Neu BCL2 (n 5) animals. Photos were taken at 200 with a Zeiss microscope. Results E13.5 Pax2 1Neu Mutant Mice Have Suboptimal Branching of the UB To illustrate the defect in UB branching in heterozygous Pax2 mutant fetal kidneys, we crossed C3H/Pax2 1Neu ( / ) mice with mice that bear a green fluorescence protein transgene that is driven by the HoxB7 promoter (C3H/HoxB7-GFP), provided by Frank Costantini (17). As seen in Figure 1, a striking reduction in UB branching is seen at E13.5 in metanephric kidneys from heterozygous Pax2 mutant mice. Targeted Overexpression of BCL2 in UB Cells of Pax2 1Neu Mutant Mice Previously, we proposed a model of renal branching morphogenesis in which the rate of UB arborization is linked to the susceptibility of UB cells to programmed cell death (12). If the antiapoptotic effects of PAX2 account for its influence on UB branching, then, we reasoned, targeted overexpression of an alternative antiapoptotic gene (BCL2) in the UB lineage should rescue the branching defect in Pax2 mutant mice. We first assessed the level of endogenous Bcl2 in fetal kidney from Pax2 mutants. Bcl2 mrna level was measured by real-time reverse transcription PCR in kidneys from four litters of E15.5 Pax2 1Neu mice (n 16) and their wild-type littermates (n 12). Similarly, endogenous Bcl2 mrna was assessed in kidneys from three litters of mutant (n 6) and wild-type (n 14) mice at P1. As seen in Figure 2, endogenous Bcl2 expression was comparable in heterozygous Pax2 mutants and controls at both stages of renal development (Figure 2). To elevate BCL2 expression in UB cells of Pax2 1Neu fetuses, Figure 1. Branching morphogenesis during fetal kidney development. At embryonic day 13,5 (E13.5), the extent of ureteric bud (UB) arborization is evident in the Hoxb7-GPF mouse. (A) The wild-type kidney has undergone several rounds of branching. (B) Conversely, a Pax2 mutant littermate has undergone fewer branching events and is hypoplastic. Bar 500 m. Figure 2. mrna expression of endogenous Bcl2 in wild-type versus Pax2 mutant kidneys. Kidneys were microdissected from E15.5 and newborn (postnatal day 1 [P1]) mice. Bcl2 mrna was measured by real-time reverse transcription PCR and standardized for a housekeeping gene ( 2-microglobulin). At E15.5 and P1, Pax2 mutant kidney Bcl2 mrna (standardized for 2-microglobulin) was normal, relative to wild-type kidney. we inserted a full-length human BCL2 cdna into an expression vector downstream of the 4-kb 5 flanking sequence of the human PAX2 gene (Figure 3A) (13). The linearized vector was

4 J Am Soc Nephrol 17: , 2006 Targeted BCL2 Expression Rescues Renal Hypoplasia 1571 BCL2 Transgene Normalizes Apoptosis in UB Cells of Pax2 1Neu ( / ) Mice To determine whether the targeted BCL2 transgene was sufficient to reverse the increased programmed cell death that was observed in UB cells of Pax2 1Neu ( / ) mutant mice, we performed TUNEL staining on P1 kidneys. In Figure 4, A and B, heterozygous mutant kidneys exhibit a 10- to 15-fold increase in TUNEL-positive UB cells compared with wild-type kidneys (Figure 4C). Mutant Pax2 1Neu ( / ) mice that bear the targeted BCL2 transgene have a striking reduction in the number of TUNEL-positive UB cells compared with mutants and are indistinguishable from wild-type mice (Figure 4D). Figure 3. Experimental overview of the generation of targeted BCL2 transgenic mice. (A) A transgene that contained human BCL2 cdna under the control of the human PAX2 promoter was linearized and inserted into pronuclei of fertilized ova. (B) The transgene was identified by PCR amplification of the IRES sequence (lanes 4, 5, 7, 8, and 9; lane 10, water control). (C) BCL2 transgene protein expression (immunohistochemistry) is seen in the UB of E15.5 kidney. (D) Minimal endogenous BCL2 protein is seen in E15.5 kidney of control littermates. (E and F) Lac Z transgene expression is restricted to UB trunk and tips but not in condensing mesenchyme of E15.5 kidney. Bar 100 m. BCL2 Transgene Restores Normal Kidney Size and Nephron Number To assess kidney size at birth, we prepared serial sagittal sections of P1 kidneys and identified the maximal cross-sections for analysis. Pax2 1Neu ( / ) kidneys (Figure 5A) had 20% smaller maximal cross-sectional area than did wild-type littermates (P 0.01; Figure 5B). In contrast, maximal cross-sectional area of P1 kidneys from Pax2 1Neu ( / ) mice that bore the targeted BCL2 transgene (Figure 5C) was significantly greater than mutants (P 0.05) and was indistinguishable from that of wild-type littermates (P 0.35). Results from four litters are summarized in Figure 5D. In other mouse models, kidney size at birth (before postnatal compensatory hypertrophy) reflects nephron number (5). We compared nephron number of P1 kidneys in two ways. First, glomeruli were counted in maximal sagittal cross-sections of kidneys from each group. Glomerular number in Pax2 1Neu ( / ) injected into the male pronucleus of fertilized C3H/C57 ova, which then were transferred to pseudopregnant CD1 female mice. Progeny were genotyped by PCR for presence of the transgene. Three founder animals were established and backcrossed for five generations into the C3H background strain. Presence of the transgene was confirmed by PCR (Figure 3B). One of the positive founders then was used for all animal crosses. Kidneys from E15.5 transgenic fetuses were studied for BCL2 protein expression by immunohistochemistry. As seen in Figure 3C, high levels of BCL2 protein are evident in the UB/collecting ducts of transgenic mice compared with the low endogenous levels in controls (Figure 3D). The presence of the transgene was confirmed further by LacZ staining of whole kidneys. Figure 3, E and F, shows positive transgenics with restricted LacZ expression in the UB structures of the developing E15.5 kidneys. Figure 4. BCL2 transgene normalizes UB cell apoptosis in Pax2 mutant mice. Apoptotic cells were identified by terminal deoxynucleotidyl-transferase mediated dutp nick end labeling (TUNEL) staining (arrows). Heterozygous Pax2 mutant mice have more numerous TUNEL-positive cells in the UB (A and B) compared with wild-type littermates (C). Targeted expression of the BCL2 transgene to UB significantly reduces the number of TUNEL-positive cells (D). Bar 100 m.

5 1572 Journal of the American Society of Nephrology J Am Soc Nephrol 17: , 2006 Figure 5. BCL2 transgene rescues kidney size at birth (P1). Maximal cross-sectional surface areas were determined for each kidney. Heterozygous Pax2 mutants (n 9; A and D) have smaller kidneys than wild-type littermates (n 12; P 0.005; B and D). Kidneys from Pax2 mutants that express the targeted BCL2 transgene (n 8; C and D) have significantly larger maximal cross-sectional area than mutant littermates (P 0.05) and are not significantly different from wild-type littermates (P 0.35). X-section, cross-section. Bar 500 m. mutants ( SEM per maximal cross-section) was only 71.5% (P 0.01) of that of wild-type littermates ( SEM per maximal cross-section; Figure 6A). By contrast, glomerular number of mutant pups that bore the BCL2 transgene ( per maximal cross-section) was significantly greater than that of mutants (P 0.01) and comparable to that of controls (P 0.52). We also estimated glomerular endowment by the physical dissector/fractionator method of unbiased stereology (5). At P1, Pax2 1Neu ( / ) mutants ( glomeruli/kidney) had only 60% as many glomeruli (P 0.004) as their wild-type littermates ( glomeruli/kidney; Figure 6B). It is interesting that the effect of the mutant Pax2 allele on glomerular number varied considerably among affected mice. Again, the number of glomeruli in kidneys from Pax2 1Neu ( / ) mutant mice that bore the BCL2 transgene ( glomeruli/ kidney) was greater than that in mutants (P 0.01) and was similar to that of wild-type littermates (P 0.09). It is interesting that kidney cross-sections of wild-type P1 mice that bore the BCL2 transgene (n 8) were 14 5% larger than that of their wild-type littermates (n 12) without the transgene (P 0.05). Glomeruli per maximal cross-section in the transgenic mice were 28 10% (P 0.05) greater than that in wild-type littermates. Figure 6. Targeted BCL2 expression restores congenital nephron number. (A) Glomeruli were counted in the maximal crosssection of each P1 kidney. Glomerular number in Pax2 1Neu ( / ) mutants was only 71.5% of that in wild-type littermates (P 0.01). Glomerular number of mutant pups that bear the BCL2 transgene was significantly greater than that in mutants without the transgene (P 0.01) and comparable to that of controls (P 0.52). (B) The number of glomeruli per kidney was estimated by the dissector/fractionator method in P1 kidneys. Pax2 1Neu ( / ) mutants had only 60% as many glomeruli as wild-type littermates (P 0.004). Glomerular number in Pax2 1Neu ( / ) mutants that bear the BCL2 transgene was significantly increased compared with that in mutants that lack the transgene (P 0.01) and was 90% of that of wild-type littermates (P 0.09). Four animals were analyzed in each group. BCL2 Transgene Rescues Renal Insufficiency at 1 Yr of Age Renal outcomes in adult Pax2 1Neu ( / ) mice have not been examined to date. However, some children with renalcoloboma syndrome may develop progressive renal insufficiency during childhood (8,7). We measured serum cystatin as a surrogate of GFR at 10 to 13 mo of age in each group of mice (Figure 7). Pax2 1Neu ( / ) mice had significantly higher serum cystatin C levels ( SEM mg/l) than that of their wild-types ( SEM mg/l; P 0.001). In contrast, mean cystatin C level in Pax2 1Neu mice that bore the BCL2 transgene ( SEM mg/l) was significantly lower than that of mutants (P 0.03) and was comparable to that of wild-type mice (P 0.08). In humans with reduced nephron endowment, compensatory glomerular hypertrophy has been well documented (6,10). We noted striking glomerular hypertrophy in Pax2 1Neu ( / ) mutants but only minimal hypertrophy in mutant mice that

6 J Am Soc Nephrol 17: , 2006 Targeted BCL2 Expression Rescues Renal Hypoplasia 1573 ( mm 3 ) was 205% that of wild-type controls ( mm 3 ; Figure 8D). Pax2 1Neu ( / ) animals that bore the BCL2 transgene had significantly less glomerular hypertrophy ( mm 3 ) compared with mutant animals (P 0.005), although glomerular volume still was slightly greater (116%) compared with wild-type mice (P 0.05). Figure 7. BCL2 normalizes renal function in adult Pax2 mutant mice. At 1 yr of age, Pax2 mutant mice (n 12) had significantly higher serum cystatin levels (P ) than wild-type littermates. In Pax2 mutant mice that bore the BCL2 transgene (n 7), serum cystatin C levels (mg/l) at 1 yr were significantly below that of mutants that lacked the transgene (n 7; P 0.05) and were similar to wild-type littermates (P 0.08). bore the BCL2 transgene (Figure 8, A through C). We measured maximal glomerular diameters and estimated mean glomerular volume in each of the three groups at 10 to 13 mo of age. Maximal glomerular volume in Pax2 1Neu ( / ) mutant kidneys Figure 8. Glomerular hypertrophy in adult Pax2 mutant mice. Mean glomerular volume was calculated from the diameters of the largest 20 glomeruli in each of four representative sections (sections 40, 90, 140, and 190) from each kidney. Representative glomeruli from wild-type (A), heterozygous Pax2 mutant (B) and Pax2 mutants that bear the BCL2 transgene (C) are displayed. The glomeruli of Pax2 mutants (n 5) are significantly larger than those of wild-type littermates (n 4; P 0.005). Pax2 1Neu BCL2 animals (n 5) had much smaller glomerular volumes than Pax2 mutants that lacked the transgene (P 0.005), although they still exhibited mild hypertrophy compared with wild-type mice (P 0.05). Bar 100 m. Discussion The factors that bring nephrogenesis to an end before birth in humans (which corresponds to approximately 1 to 2 wk after birth in rodents) are unknown, but once the window of opportunity closes, final nephron number has been irreversibly set. Conceivably, nephron number is influenced by the pool size of blastemal stem cells. However, our observations suggest that another limiting factor might be the rate at which the UB arborizes during fetal life. Whatever the mechanism, neonates who are born with low nephron number may develop progressive renal insufficiency as somatic growth produces an increasing discrepancy between workload and excretory capacity. Approximately 40% of pediatric ESRD is due to some form of congenital hypoplasia ( dysplasia) (18,19). Even when postnatal compensatory hypertrophy achieves temporary adaptation to postnatal life, there is growing evidence that subtle renal hypoplasia, within the spectrum of normal, may be a risk factor for susceptibility to acquired renal disease later in life. Furthermore, the mechanisms that support compensatory hypertrophy may be central to the pathogenesis of essential hypertension (6). In a previous study, we proposed a model in which the rate of UB branching is linked directly to the susceptibility of UB cells to apoptosis (13). This idea was based on two main observations: (1) That heterozygous Pax2 fetal mice have a selective increase in the percentage of TUNEL-positive UB cells and (2) that the collecting system has undergone fewer branching events in mutants compared with wild-type littermates by E15. However, it has not yet been shown that early retardation of UB branching translates into a congenital nephron deficit. Furthermore, the functions of PAX2 during development are multiplex and might influence branching nephrogenesis through some other pathway. PAX2 affects cell fate during descent of the nephric duct (20) and has been shown to activate Gdnf in the metanephric mesenchyme (21). In this study, we tested the hypothesis that any change in the balance of factors that determine programmed cell death in UB cells will modify the extent of arborization during fetal life and determine nephron endowment for life. The mechanism by which PAX2 opposes the apoptotic pathway is yet unknown. PAX2 effects the expression of genes in the metanephric mesenchyme such as Gdnf (21) and Wnt4 (22). However, the observation that apoptosis of cultured inner medullary collecting duct cells is strongly suppressed by PAX2 suggests a transcriptional target in cells of the UB lineage as well (23). Our previous studies indicated that increased apoptosis is restricted to the UB in Pax2 1Neu mice (12). In this study, the BCL2 transgene was selectively targeted to the UB by a 4-kb PAX2 promoter. Nowhere within the first 10 kb of the 5

7 1574 Journal of the American Society of Nephrology J Am Soc Nephrol 17: , 2006 flanking sequence of the Pax2 gene could Kuschert et al. (24) locate the Pax2 promoter elements that control PAX2 expression in the mesenchyme. Although we cannot rule out the possibility that PAX2 affects UB apoptosis through an indirect mechanism such as GDNF release from the mesenchyme (25), our results suggest an antiapoptotic PAX2 mechanism that is intrinsic to the UB. Because many different proapoptotic stimuli converge on mitochondrial release of cytochrome c to initiate the caspase cascade, we measured fetal kidney expression of BCL2, a gatekeeper of cytochrome c release through the inner mitochondrial membrane. Bcl2 mrna levels were entirely normal in Pax2 mutant mice. Although this excludes Bcl2 as a likely gene target of Pax2, it nevertheless provides an interesting experimental strategy by which to manipulate the balance of apoptotic factors in UB cells without compromising other PAX2 developmental functions. BCL2 is widely expressed during fetal development, but its expression becomes highly restricted as organs mature (26). At E12, Bcl2 is normally expressed both in the UB and in the condensing metanephric mesenchyme (27). In our transgenic mice, BCL2 was targeted correctly to the UB by the 4-kb human PAX2 promoter and was expressed well above endogenous levels in fetal kidneys. This was sufficient to normalize apoptotic rates in the UB and restore maximal cross-sectional area of newborn kidneys to the range of that of wild-type littermates. Rescue of renal hypoplasia correlated with normalization of newborn nephron number. These observations clearly demonstrate a direct link between the extent of UB arborization at E15.5 and congenital nephron number. We did not initially anticipate an effect of the BCL2 transgene on wild-type mice, so detailed stereologic assessment was not performed on these animals. However, we noted that P1 kidney size and glomerular number per maximal cross-section were significantly increased in the wild-type mice that bore a BCL2 transgene versus wild-type controls. Our earlier observations identified a low but measurable number of TUNEL-positive cells in the normal UB, which could account for the transgene effect on wild-type kidneys (12). The observations in BCL2 transgenic mice are consistent with our proposed model of branching morphogenesis. If increased apoptosis slows the rate at which a branch emerges from the putative inhibitory field, then suppression of basal apoptosis should lead to greaterthan-normal nephron number. Although the Pax2 mutant mouse has been used as an important model of congenital renal hypoplasia, long-term studies of renal outcome have not been reported. In this study, we examined renal outcome in Pax2 heterozygotes at 1 yr. Pax2 mutant mice had significantly larger glomerular volumes than wild-type littermates, and serum cystatin C levels were increased, although no significant increase in glomerulosclerosis was seen. Therefore, mutant mice exhibited incomplete compensation for the congenital nephron deficit. Targeted overexpression of BCL2 during the period of nephrogenesis was sufficient to restore nephron number to 90% of that of controls and normalize adult serum cystatin levels, although there still was some residual glomerular hypertrophy. These data strongly support a model of renal branching morphogenesis in which susceptibility of UB cells to programmed cell death influences congenital nephron number. By correcting the deficit of UB cell antiapoptotic factors that are imposed by a Pax2 gene mutation, we show that nephron number and long-term renal function can be rescued. The renal hypoplastic effects of Pax2 haploinsufficiency are seen in humans and in several strains of mice, suggesting that the mechanism is highly conserved and does not depend on unidentified modifier genes. It is conceivable, therefore, that dysregulation of programmed cell death in fetal kidney could account for suboptimal nephron endowment in humans, as well. Acknowledgments This study was supported by operating grants from the Canadian Institutes of Health Research (MOP 12954) and from the Cancer Society of New Zealand (CSNZ01/3). A.D. is a recipient of a Canadian Institutes of Health Research doctoral studentship award. P.G. is the recipient of a James McGill Research Chair. Preliminary data were presented at the meeting of the American Society of Nephrology, November 14, 2003, San Diego, CA. References 1. Nyengaard JR, Bendtsen TF: Glomerular number and size in relation to age, kidney weight, and body surface in normal man. Anat Rec 232: , Simons JL, Provoost AP, Anderson S, Rennke HG, Troy JL, Brenner BM: Modulation of glomerular hypertension defines susceptibility to progressive glomerular injury. Kidney Int 46: , Brenner BM, Milford EL: Nephron underdosing: A programmed cause of chronic renal allograft failure. Am J Kidney Dis 21: 66 72, Brenner BM, Chertow GM: Congenital oligonephropathy and the etiology of adult hypertension and progressive renal injury. Am J Kidney Dis 23: , Cullen-McEwen LA, Kett MM, Dowling J, Anderson WP, Bertram JF: Nephron number, renal function, and arterial pressure in aged GDNF heterozygous mice. Hypertension 41: , Keller G, Zimmer G, Mall G, Ritz E, Amann K: Nephron number in patients with primary hypertension. N Engl J Med 348: , Sanyanusin P, McNoe LA, Sullivan MJ, Weaver RG, Eccles MR: Mutation of PAX2 in two siblings with renalcoloboma syndrome. Hum Mol Genet 4: , Sanyanusin P, Schimmenti LA, McNoe LA, Ward TA, Pierpont ME, Sullivan MJ, Dobyns WB, Eccles MR: Mutation of the PAX2 gene in a family with optic nerve colobomas, renal anomalies and vesicoureteral reflux. Nat Genet 9: , Dressler GR, Wilkinson JE, Rothenpieler UW, Patterson LT, Williams-Simons L, Westphal H: Deregulation of Pax-2 expression in transgenic mice generates severe kidney abnormalities. Nature 362: 65 67, Salomon R, Tellier AL, Attie-Bitach T, Amiel J, Vekemans M, Lyonnet S, Dureau P, Niaudet P, Gubler MC, Broyer M: PAX2 mutations in oligomeganephronia. Kidney Int 59: , Favor J, Sandulache R, Neuhauser-Klaus A, Pretsch W,

8 J Am Soc Nephrol 17: , 2006 Targeted BCL2 Expression Rescues Renal Hypoplasia 1575 Chatterjee B, Senft E, Wurst W, Blanquet V, Grimes P, Sporle R, Schughart K: The mouse Pax2(1Neu) mutation is identical to a human PAX2 mutation in a family with renal-coloboma syndrome and results in developmental defects of the brain, ear, eye, and kidney. Proc Natl Acad Sci USA93: , Porteous S, Torban E, Cho NP, Cunliffe H, Chua L, McNoe L, Ward T, Souza C, Gus P, Giugliani R, Sato T, Yun K, Favor J, Sicotte M, Goodyer P, Eccles M: Primary renal hypoplasia in humans and mice with PAX2 mutations: Evidence of increased apoptosis in fetal kidneys of Pax2(1Neu) / mutant mice. Hum Mol Genet 9: 1 11, Dziarmaga A, Clark P, Stayner C, Julien JP, Torban E, Goodyer P, Eccles M: Ureteric bud apoptosis and renal hypoplasia in transgenic PAX2-Bax fetal mice mimics the renal-coloboma syndrome. J Am Soc Nephrol 14: , Bertram JF: Counting in the kidney. Kidney Int 59: , Bertram JF, Nurcombe V: Counting cells with the new stereology. Trends Cell Biol 2: , Filler G, Priem F, Vollmer I, Gellermann J, Jung K: Diagnostic sensitivity of serum cystatin for impaired glomerular filtration rate. Pediatr Nephrol 13: , Srinivas S, Goldberg MR, Watanabe T, D Agati V, al- Awqati Q, Costantini F: Expression of green fluorescent protein in the ureteric bud of transgenic mice: A new tool for the analysis of ureteric bud morphogenesis. Dev Genet 24: , Ardissino G, Dacco V, Testa S, Bonaudo R, Claris-Appiani A, Taioli E, Marra G, Edefonti A, Sereni F: Epidemiology of chronic renal failure in children: Data from the ItalKid project. Pediatrics 111: e382 e387, Tejani AH, Sullivan EK, Harmon WE, Fine RN, Kohaut E, Emmett L, Alexander SR: Pediatric renal transplantation The NAPRTCS experience. Clin Transpl , Bouchard M, Souabni A, Mandler M, Neubuser A, Busslinger M: Nephric lineage specification by Pax2 and Pax8. Genes Dev 16: , Brophy PD, Ostrom L, Lang KM, Dressler GR: Regulation of ureteric bud outgrowth by Pax2-dependent activation of the glial derived neurotrophic factor gene. Development 128: , Torban E, Dziarmaga A, Iglesias D, Chu LL, Vassilieva T, Little M, Eccles M, Discenza M, Pelletier J, Goodyer P: PAX2 activates Wnt4 expression during mammalian kidney development. J Biol Chem December 19, 2005 [epub ahead of print] 23. Torban E, Eccles MR, Favor J, Goodyer PR: PAX2 suppresses apoptosis in renal collecting duct cells. Am J Pathol 157: , Kuschert S, Rowitch DH, Haenig B, McMahon AP, Kispert A: Characterization of Pax-2 regulatory sequences that direct transgene expression in the Wolffian duct and its derivatives. Dev Biol 229: , Towers PR, Woolf AS, Hardman P: Glial cell line-derived neurotrophic factor stimulates ureteric bud outgrowth and enhances survival of ureteric bud cells in vitro. Exp Nephrol 6: , Veis DJ, Sorenson CM, Shutter JR, Korsmeyer SJ: Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75: , Novack DV, Korsmeyer SJ: Bcl-2 protein expression during murine development. Am J Pathol 145: 61 73, 1994